A method of encoding a video includes: splitting a picture into a maximum coding unit; for the maximum coding unit, determining coding units having a tree structure including coding units of coded depths and determining encoding modes for the coding units of the coded depths by performing encoding based on coding units according to depths, the coding units according to depths obtained by hierarchically splitting the maximum coding unit as a depth deepens; and outputting information about a maximum coding unit size and, for the maximum coding unit, information indicating an order of split information and skip mode information which is selectively determined for the coding units according to depths, information about the encoding modes for the coding units of the coded depths including the split information and the skip mode information which are arranged according to the order, and encoded video data.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method of decoding a video, the method comprising: receiving a bitstream of encoded video; extracting, from the bitstream, split information of a coding unit in a maximum coding unit of a picture; when the split information indicates a split for a current depth, splitting the coding unit of the current depth, independently from neighboring coding units, into coding units of a lower depth; and when the split information indicates a non-split for the current depth, parsing a skip mode information of the coding unit of the current depth indicating whether a prediction mode of the coding unit is a skip mode, wherein, when skip information for the coding unit of the current depth indicates a skip mode, the coding unit of the current depth is decoded, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or by inferring motion information of the coding unit from motion information of the adjacent prediction unit, wherein, when the skip information for the coding unit of the current depth does not indicate the skip mode, the coding unit of the current depth is decoded by determining at least one partition from the coding unit of the current depth, performing prediction on the coding unit of the current depth based on the at least one prediction unit in a prediction mode indicated by prediction mode information obtained from the bitstream, determining at least one transformation unit obtained from the coding unit of the current depth and performing inverse transformation on the coding unit of the current depth based on the at least one transformation unit, and wherein the coding unit of the current depth is split into the coding units of the lower depth by halving a height and a width of the coding unit of the current depth when the split information indicates a split for a current depth.
2. A method of encoding a video, the method comprising: generating split information of a coding unit of a current depth in a maximum coding unit of a picture; when the coding unit of the current depth is not split into coding units of a lower depth, generating skip information of the coding unit of the current depth; and when prediction is performed on the coding unit of the current depth not in a skip mode, generating information about a prediction mode indicating one of an intra mode and an inter mode and generating information about a partition type of the coding unit of the current depth, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, when the coding unit of the current depth is not split into the coding units of the lower depth, at least one partition is determined from the coding unit of the current depth and a prediction is performed on the coding unit of the current depth based on the at least one partition, when prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit, and when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth, the information about the partition type indicates a size of the at least one partition, at least one transformation unit is obtained from the coding unit of the current depth and transformation is performed on the coding unit of the current depth based on the at least one transformation unit.
3. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising: split information of a coding unit of a current depth in a maximum coding unit of a picture; and skip information of the coding unit of the current depth to indicate whether a prediction is performed on the coding unit of the current depth in a skip mode, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, when the coding unit of the current depth is not split into the coding units of the lower depth, the skip information of the coding unit of the current depth is generated, when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth, a prediction is performed on the coding unit of the current depth based on the at least one partition, at least one transformation unit is obtained from the coding unit of the current depth and transformation is performed on the coding unit of the current depth based on the at least one transformation unit, when the prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed in the skip mode, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit.
4. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising: split information of a coding unit of a current depth in a maximum coding unit of a picture; skip information of the coding unit of the current depth to indicate whether a prediction is performed on the coding unit of the current depth in a skip mode; and information about a prediction mode of the coding unit of the current depth to indicate one of an intra mode and an inter mode, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, when the coding unit of the current depth is not split into the coding units of the lower depth, the skip information of the coding unit of the current depth is generated, when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth, a prediction is performed on the coding unit of the current depth based on the at least one partition in a prediction mode indicated by the information about a prediction mode, at least one transformation unit is obtained from the coding unit of the current depth and transformation is performed on the coding unit of the current depth based on the at least one transformation unit, when the prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed in the skip mode based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit.
5. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising: split information of a coding unit of a current depth in a maximum coding unit of a picture; skip information of the coding unit of the current depth to indicate whether a prediction is performed on the coding unit of the current depth in a skip mode; information about a prediction mode of the coding unit of the current depth to indicate one of an intra mode and an inter mode; and information about a partition type to indicate a size of at least one partition determined from the coding unit of the current depth, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, and when the coding unit of the current depth is not split into the coding units of the lower depth, the skip information of the coding unit of the current depth is generated, when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth based on the information about a partition type, a prediction is performed on the coding unit of the current depth based on the at least one partition in a prediction mode indicated by the information about a prediction mode, at least one transformation unit is obtained from the coding unit of the current depth and transformation is performed on the coding unit of the current depth based on the at least one transformation unit, when the prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed in the skip mode, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit.
6. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising: split information of a coding unit of a current depth in a maximum coding unit of a picture; skip information of the coding unit of the current depth to indicate whether a prediction is performed on the coding unit of the current depth in a skip mode; information about a prediction mode of the coding unit of the current depth to indicate one of an intra mode and an inter mode; and information about a partition type to indicate a size of at least one partition determined from the coding unit of the current depth, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, when the coding unit of the current depth is not split into the coding units of the lower depth, the skip information of the coding unit of the current depth is generated, when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth based on the information about a partition type, a prediction is performed on the coding unit of the current depth based on the at least one partition in a prediction mode indicated by the information about a prediction mode, at least one transformation unit is obtained from the coding unit of the current depth and transformation is performed on the coding unit of the current depth based on the at least one transformation unit, when the prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed in the skip mode, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit, the information about the partition type indicates one of symmetric partition types and asymmetric partition types.
7. A non-transitory computer-readable storage medium storing a bitstream, the bitstream comprising: split information of a coding unit of a current depth in a maximum coding unit of a picture; skip information of the coding unit of the current depth to indicate whether a prediction is performed on the coding unit of the current depth in a skip mode; information about a prediction mode of the coding unit of the current depth to indicate one of an intra mode and an inter mode; information about a partition type to indicate a size of at least one partition determined from the coding unit of the current depth; and information about a size of a transformation unit to indicate a size of at least one transformation unit determined from the coding unit of the current depth, wherein: when the coding unit of the current depth is split into the coding units of the lower depth, the coding unit of the current depth is split independently from neighboring coding units, into the coding units of the lower depth, when the coding unit of the current depth is not split into the coding units of the lower depth, the skip information of the coding unit of the current depth is generated, when the prediction is performed on the coding unit of the current depth not in the skip mode, at least one partition is determined from the coding unit of the current depth based on the information about a partition type, a prediction is performed on the coding unit of the current depth based on the at least one partition in a prediction mode indicated by the information about a prediction mode, at least one transformation unit is obtained from the coding unit of the current depth based on the information about a size of a transformation unit and transformation is performed on the coding unit of the current depth based on the at least one transformation unit, when the prediction is performed on the coding unit of the current depth in the skip mode, the prediction is performed in the skip mode, based on a partition having a size equal to a size of the coding unit of the current depth, by borrowing motion information of a prediction unit adjacent to the coding unit, or inferring motion information of the coding unit from motion information of the adjacent prediction unit, the information about the partition type indicates one of symmetric partition types and asymmetric partition types.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
January 27, 2020
September 21, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.